ANALYTICAL
65, 225-23 1 ( 1975)
BIOCHEMISTRY
A Simplified
Microassay Using
of DNA
Ethidium
and
RNA
Bromide
G. J. BOER Netherlands
Central Institute for Brain Research, IJdijk 28, Amsterdam-O
Received August 26, 1974; accepted December 4. I974 The ethidium bromide method of Karsten and Wollenberger (3) has been modified allowing determination of DNA and RNA in small samples of tissue homogenates (IO-30 pg wet weight) or in small pieces of frozen-dried tissue sections (2-5 pg dry weight). The detection limit for DNA ranges between 5 and 10 ng, depending on the origin of the tissue, and for RNA is twice as high. The present micromethod offers a good fluorimetric alternative for the more laborious diaminobenzoic acid DNA assay of Kissane and Robins (4).
Measurement of as little as 2.4 ng DNA was described in 1958 by Kissane and Robins (4). This procedure, however, requires a considerable amount of manipulation. A simple fluorimetric method allowing estimation of both DNA and RNA in solution has been described (5) and was adapted for homogenates (3). This method is based on the enhancement of the fluorescence of ethidium bromide after complexing with nucleic acids (6). However, the detection limit for tissue DNA in the procedure utilizing ethidium bromide is about 50 ng DNA. During our study of small samples of rat neurohypophysis (2) a need was felt for a simple microdetermination of DNA. Therefore, the ethidium bromide method was adapted for DNA assay in very small samples. The present paper describes this method using microdissected pieces of frozen-dried tissue (7) and small volumes of homogenate from both neurohypophysis and liver. MATERIALS AND METHODS
Chemicals. All chemicals were from Merck, except for pronase (free of nucleases, 120,000 PUK/g, Calbiochem); DNAase (free of RNAase, 2880 Kunitz units/mg, DN-C) and RNAase (100 Kunitz units/mg, Type I-A) (both Sigma), RNA (from yeast, Worthington), DNA (from calf thymus, Sigma type V), ethidium bromide (Boots Pure Drug Company Ltd.) and 3,5-diaminobenzoic acid (Fluka). Preparation of the tissue. The pituitary and liver of 180- to 250-g male Wistar rats, sacrificed by decapitation, were isolated and either placed in ice-cold 0.9% NaCl before preparing homogenates (procedure A) or 225 Copyright @ 1975 by Academic Press, Inc. All rights of reproduction in any form reserved.
226
G. .I. BOER
frozen directly into liquid nitrogen for cryostat sectioning and subsequent freeze-drying (procedure B). Procedure A. Homogenates were made in 50- to 150~~1 phosphatebuffered saline (see below) from small tissue samples (1 S-30 mg) using a pestle homogenizer (8) (1000 rpm). Neurohypophyses were separated from the anterior and intermediate lobe under the stereomicroscope and three were pooled (about 1.8 mg) for homogenization. Procedure B. Frozen tissue was sectioned (16 pm) and subsequently lyophilized according to the procedures of Lowry (7). Small pieces (2-5 pg dry weight) were dissected out and weighed on quartz-fiber balances (7). Determination of DNA. Small volumes of homogenate (IO-40 pg fresh weight in 5 ~1) or a piece of frozen-dried tissue were incubated for 30 min at 37°C in a final volume of 20 ,ul phosphate-buffered saline (0.0 1% CaC&, 0.02% KCl, 0.02% KH2P04, 0.01% MgCl, . 6H10, 0.8% NaCl, 0.115% Na,HPO,, adjusted to pH 7.5 with NaOH; ref. 3), containing 0.1 mg/ml pronase and 50 pg/ml RNAase. The medium was placed into a Corning tube (i.d. 4 mm) and the tube capped with Parafilm. After incubation 75 ~1 phosphate-buffered saline was added and background fluorescence readings were made on a Vitatron fluorometer. The fluorescence was measured again after addition of ethidium bromide (5 ~1 50 @g/ml). The excitation (360 nm) and emission wavelengths (590 nm) of ethidium bromide and its complex with nucleic acids (5) were isolated using Corning No. 5860 as primary filter and Jena No. OG 1 as secondary “cut-off” filter. The fluorescence can be measured within 1 min after addition and remains stable for at least 3 hr. Blanks and standards (lo-100 ng DNA) were included in the whole procedure. Ethidium bromide and DNA standard were dissolved in phosphate-buffered saline. A DNA stock solution (23 mg/ml) was prepared in 0.1 M ammonia and kept at 4°C. Determination of RNA. If both RNA and DNA content had to be measured, tissue samples were incubated as described above but also without addition of RNAase. RNA standards (20-300 ng) could be introduced, but the alternative for RNA standards, i.e., the use of DNA standards only and using a factor 0.46 as quotient of RNA/DNA fluorescence (c$ ref. 3 and 5), indeed appeared to give similar results. A RNA stock solution (23 mglml) was prepared in phosphate-buffered saline and kept at 4°C. Freeze-drying. In order to examine the effect of freeze-drying, volumes of 20 ~1 homogenate were lyophilized (0.03 mmHg for 5 hr at -5°C) and redissolved in the same volume of water. DNA content was determined according to the present method as well as by following the procedures of Kissane and Robins (4).
ETHIDIUM
BROMIDE
ASSAY
OF
DNA
AND
227
RNA
Statistics. The significance of differences between the values was calculated using the Student t test. A value for p < 0.05 was considered to be statistically significant. RESULTS
Proportionality. Using 2.5 @g/ml ethidium bromide, the fluorescence shows a linear relationship with the DNA content between 2.5 and 100 ng in the 100~~51final volume. The enhanced fluorescence of 2.5 ng DNA differs significantly (10%) from the ethidium bromide blank (0.025 < p < 0.05; II = 6). For RNA the same appears to be valuable between 5 and 200 ng, i.e., twice the amount of DNA. The recovery of DNA-ethidium bromide fluorescence for contents between 10 and 200 ng appeared to be completely after pronase (0.4 mglmi) treatment (102%; SEM 1.6%; n = 10). Pronase treatmerzt. In order to increase the sensitivity of the ethidium bromide DNA assay, the high background fluorescence due to the tissue (see below), pronase and RNAase had to be suppressed. Reduction of pronase and RNAase background fluorescence was obtained by incubation in a small volume which can be diluted five times prior to background fluorescence reading. Different pronase concentrations were tested for their induction of ethidium bromide fluorescence of tissue samples. For homogenate (Table 1) as well as for pieces of frozen-dried sections of the rat neurohypophysis a broad optimum of fluorescence was found between 0.1 and TABLE ENHANCEMENT
OF ETHIDIUM LOBE
BROMIDE
HOMOGENATE
1 FLUORESCENCE
OF RAT
NEURAL
BY PRONASE”
Pronase
concentration
Fluorescenceh
(mgimlf 0
0.004 0.01 0.04 0.1 0.2 0.4 4 8 20
f%‘o)
100 +- 5.4 99 160 208 22.5 230 223 141 110 68
it 6.9 2 4.9 4 4.9 t 4.1 + 6.6 zt 2.3 2.z 4.8 it 3.6 rt 5.2
n Procedure as described in Material and Methods section using about 3Opg fresh weight of rat neural lobe tissue. b Values are given in percentage (?SEM) of the total fluorescence without pronase treatment and are the mean of three determinations.
228
G. J. BOER TABLE
INFLUENCE WITH
2
OF RNAase IN THE PRESENCE OF PRONASE ON THE FLUORESCENCE ETHIDIIJM BROMIDE OF RNA AND DNA ADDED TO A SAMPLE OF RAT NEURAL LOBE HOMOGENATE”
Fluorescence (%‘oy’
Homogenate +200 ng RNA + 100 ng DNA
+ RNAase
-RNAase
100 f 3.6 102 + 7.2 312 ” 11
369 k 13 382 k 20
163 r 7.2
a Procedure as described in Material and Methods section using about 30 Fg fresh weight rat neural lobe and 5 pg/ml ethidium bromide for final fluorescence reading. b Values are given in percentage (?SEM) of the DNA-ethidium bromide fluorescence of the homogenate and are the mean of three determinations.
0.4 mglml (0.5 < p < 0.6) after 30-min incubation at 37°C. Below 0.1 mg/ml the maximum was not reached; at 20 mglml the fluorescence became even lower than without incubation. The concentration of 0.1 mg/ml pronase was chosen in the final procedure, giving an autofluorescence of about 6% of the ethidium bromide blank fluorescence. Incubation with nucleuses. In the presence of pronase RNAase should break down all RNA present in the tissue sample without attacking DNA. In our procedure the fluorescence of 200 ng RNA, added to a homogenate sample, disappeared completely with the RNAase concentration used (50 pglml), while the fluorescence due to 100 ng DNA added remained present (Table 2). Both nucleic acids were added in amounts that were expected to be present in the 30 pg of tissue maximally used for the micro DNA assay. DNAase (50 pglml) degrades 100 ng DNA totally, independent of the presence of homogenate and pronase. Since no residual fluorescence was observed, DNAase treatment had not to be incorporated in the microassay. Linearity with tissue samples. The linear relation between ethidium bromide fluorescence and sample size is shown for liver homogenate in Fig. 1. Since this linearity is also present in the absence of RNAase, both DNA and RNA can be determined. Background JEuorescence. The background fluorescence for liver is very high (about the value of the enhancement in fluorescence caused by the DNA in the sample), which reduces the sensitivity of the assay seriously (see Discussion). Nevertheless as little as 5 pg wet weight of liver tissue appeared to be enough to give reproducible fluorescence increases. For this weight (the lowest value of Fig. 1) the enhanced fluorescence for the DNA present is about 10% of the total fluorescence. This differs significantly (20%) from the ethidium bromide blank
ETHIDIUM
BROMIDE
229
ASSAY OF DNA AND RNA
FIG. 1. Linear relationship between the amount of rat liver tissue and the enhancement of the ethidium bromide fluorescence by its DNA content (lower line) and by its total nucleic acid content (upper line). Each point indicates the mean of three determinations.
fluorescence (0.025 < p =C0.05; n = 3) and corresponds to about 11 ng of DNA. The background fluorescence of liver expressed per unit of wet weight was about 5 times as high as for the neurohypophysis. For both liver (Fig. 1) and neurohypophysis the tissue background fluorescence did not interfere with the linearity of fluorescence and sample size. Freeze-drying of the homogenates had no apparent effect on the DNA TABLE DNA
3
AND RNA CONTENT OF RAT NEURAL LOBE AND DIFFERENT PRETREATMENTS OF THE TISSUE Two DIFFERENT MICROASSAYS
LIVER MEASURED AFTER AND USING
DNA content? Present micromethod Neural lobe Homogenateb Fresh Frozen-dried
1.16 t 0.04 1.22 2 0.07
Frozen-dried section samples* 9.48 k 1.3
Liver 2.76 k 0.09 2.48 2 0.14 13.3 + 0.9
Method of Kissane and Robins Neural lobe 1.08 ? 0.05 1.30 k 0.20 10.9 + 0.7
Liver 2.41 t 0.27 2.67 5 0.14 14.6 + 1.0
RNA content” Neural lobe
Liver
1.84 n.d.c
6.52 n.d.
12.5
44.4
a Values expressed as ngIpg fresh weight and ngipg dry weight for homogenates and section samples, respectively; each value is the mean (SEM) of three determinations for the homogenates and of five for the section samples. * Homogenate of neurohypophysis originates from three animals from which one was used for preparing liver homogenate; frozen-dried section samples originate from one single animal. C Not determined.
230
G. J. BOER
content measured with the ethidium bromide procedure as well as with the diaminobenzoic acid procedure (Table 3; p-values always above 0.30). Comparison with the diaminobenzoic acid method. After carrying out all pretreatments used for rat neurohypophysis and liver, both the present microassay and the assay with diaminobenzoic acid resulted in similar data for the DNA content (Table 3; p-values always above 0.20). The tissue background fluorescence of the present assay, however, lowered the sensitivity of DNA measurement in such a way that three times more tissue DNA is necessary to give the same increment in fluorescence compared to the blank as with the diaminobenzoic acid method (4). DISCUSSION
The proportionality of DNA and RNA with their ethidium bromide fluorescence in the nanogram range and the nucleotide differentiation using specific nucleases (Table 2) show that the DNA and RNA measurement with ethidium bromide as described by Le Pecq and Paoletti (5) is applicable also at the microlevel. As little as 2.5 ng DNA or 5 ng RNA in solution can be reproducibly detected. Karsten and Wollenberger (3) introduced pronase treatment of tissue homogenate and surface fluorometry which also allows a quantitative reaction of tissue nucleic acids with ethidium bromide. Pronase treatment was also applied within the microprocedure, although sonication of tissue homogenate might be considered as an alternative (1). A higher pronase concentration, however, appeared to be necessary for maximal increase of the fluorescence. For the DNA assay also the RNAase activity had to be raised in order to break down all the RNA expected. Normal fluorometry instead of surface fluorometry could be applied in the present microassay because a linear relationship exists between RNA- and DNA-ethidium bromide fluorescence and sample size with this simpler procedure (Fig. 1). If tissue DNA and RNA content were measured with the present microassay, the sensitivity limit was lowered two to three times compared to the assay for dissolved nucleic acids, mainly by the tissue-induced background fluorescence (which, moreover, appeared to be dependent on the origin of the tissue). The minimum of handling, however, makes the ethidium bromide method favorable if more than about 10 ng of native DNA is available for an assay. Moreover, the ethidium bromide method also introduces a micro RNA determination in small tissue samples. Results of DNA estimations in rat neural lobe and liver homogenates show good agreement with data obtained following the Kissane and Robins (4) microprocedure (Table 3, first line) and are fully comparable
ETHIDIUM
BROMIDE
ASSAY OF DNA AND RNA
231
with available literature values (3,9). Also the presented RNA values are fully in agreement with the data of these authors. Freezing and thawing of tissue did not influence nucleic acid-enhanced ethidium bromide fluorescence (3), while also freeze-drying of tissue homogenates had no influence on the DNA assay (Table 3, second line). Therefore, the present method seems also valuable for small frozendried tissue samples as are used in microbiochemistry (7). Using such samples prepared from rat neural lobe and liver, indeed, with both micromethods similar data were obtained for the DNA content (Table 3, third line). The reproducibility of the present microassay lies in the same order as for the diaminobenzoic acid method (Tables 2 and 3). The present method has already been used for several months for quantitative determination of the accumulation of neurohypophyseal glial cells (pituicytes) of the rat along different micro density gradients (Ficoll and Ludox) after centrifugation. Good recoveries were obtained. If necessary, the procedure can also be scaled up two or three times if more sample is available, thus avoiding the errors inherent in handling small volumes. ACKNOWLEDGMENTS I am grateful to Dr. D. F. Swaab for his critical remarks in the present study, and thank Mr. B. Fisser, Miss A. Nolten, Mrs. C. van Rheenen. and Miss A. van der Velden for their successive assistance and Prof. Dr. J. Ariens Kappers and Prof. Dr. J. M. Tager for their revision of the manuscript.
REFERENCES 1. Blackburn. M. J., Andrews. T. M., and Watts, R. W. E. (1973) Anal. Biochem. 51, l-10. 2. Boer, G. J., and Jongkind, J. F. (1974) J. Neurochem. 22, 965-970. 3. Karsten. U., and Wollenberger, A. (1972) Amd. Biochem. 46, 135-148. 4. Kissane, J. M., and Robins, E. ( 1958) J. Biol. Chem. 233, 184- 188. 5. Le Pecq. J.-B., and Paoletti, C. (1966) And. Biochem. 17, 100-107. 6. Le Pecq, J.-B., and Paoletti, C. (1967) J. Mol. Biol. 27, 87-106. 7. Lowry, 0. H. (1953) J. Histochem. Cytochem. 1, 420-428. 8. Neuhoff. V. (1973) Micromethods in Molecular Biology, p. 399. Springer-Verlag, Berlin. 9. Sachs, H., Saito, S., and Sunde, D. (1971) in Memoirs of the Society for Endocrinology (Heller. H., and Lederis, K., eds.), p. 325. Cambridge Univ. Press, London,
65, 225-23 1 ( 1975)
BIOCHEMISTRY
A Simplified
Microassay Using
of DNA
Ethidium
and
RNA
Bromide
G. J. BOER Netherlands
Central Institute for Brain Research, IJdijk 28, Amsterdam-O
Received August 26, 1974; accepted December 4. I974 The ethidium bromide method of Karsten and Wollenberger (3) has been modified allowing determination of DNA and RNA in small samples of tissue homogenates (IO-30 pg wet weight) or in small pieces of frozen-dried tissue sections (2-5 pg dry weight). The detection limit for DNA ranges between 5 and 10 ng, depending on the origin of the tissue, and for RNA is twice as high. The present micromethod offers a good fluorimetric alternative for the more laborious diaminobenzoic acid DNA assay of Kissane and Robins (4).
Measurement of as little as 2.4 ng DNA was described in 1958 by Kissane and Robins (4). This procedure, however, requires a considerable amount of manipulation. A simple fluorimetric method allowing estimation of both DNA and RNA in solution has been described (5) and was adapted for homogenates (3). This method is based on the enhancement of the fluorescence of ethidium bromide after complexing with nucleic acids (6). However, the detection limit for tissue DNA in the procedure utilizing ethidium bromide is about 50 ng DNA. During our study of small samples of rat neurohypophysis (2) a need was felt for a simple microdetermination of DNA. Therefore, the ethidium bromide method was adapted for DNA assay in very small samples. The present paper describes this method using microdissected pieces of frozen-dried tissue (7) and small volumes of homogenate from both neurohypophysis and liver. MATERIALS AND METHODS
Chemicals. All chemicals were from Merck, except for pronase (free of nucleases, 120,000 PUK/g, Calbiochem); DNAase (free of RNAase, 2880 Kunitz units/mg, DN-C) and RNAase (100 Kunitz units/mg, Type I-A) (both Sigma), RNA (from yeast, Worthington), DNA (from calf thymus, Sigma type V), ethidium bromide (Boots Pure Drug Company Ltd.) and 3,5-diaminobenzoic acid (Fluka). Preparation of the tissue. The pituitary and liver of 180- to 250-g male Wistar rats, sacrificed by decapitation, were isolated and either placed in ice-cold 0.9% NaCl before preparing homogenates (procedure A) or 225 Copyright @ 1975 by Academic Press, Inc. All rights of reproduction in any form reserved.
226
G. .I. BOER
frozen directly into liquid nitrogen for cryostat sectioning and subsequent freeze-drying (procedure B). Procedure A. Homogenates were made in 50- to 150~~1 phosphatebuffered saline (see below) from small tissue samples (1 S-30 mg) using a pestle homogenizer (8) (1000 rpm). Neurohypophyses were separated from the anterior and intermediate lobe under the stereomicroscope and three were pooled (about 1.8 mg) for homogenization. Procedure B. Frozen tissue was sectioned (16 pm) and subsequently lyophilized according to the procedures of Lowry (7). Small pieces (2-5 pg dry weight) were dissected out and weighed on quartz-fiber balances (7). Determination of DNA. Small volumes of homogenate (IO-40 pg fresh weight in 5 ~1) or a piece of frozen-dried tissue were incubated for 30 min at 37°C in a final volume of 20 ,ul phosphate-buffered saline (0.0 1% CaC&, 0.02% KCl, 0.02% KH2P04, 0.01% MgCl, . 6H10, 0.8% NaCl, 0.115% Na,HPO,, adjusted to pH 7.5 with NaOH; ref. 3), containing 0.1 mg/ml pronase and 50 pg/ml RNAase. The medium was placed into a Corning tube (i.d. 4 mm) and the tube capped with Parafilm. After incubation 75 ~1 phosphate-buffered saline was added and background fluorescence readings were made on a Vitatron fluorometer. The fluorescence was measured again after addition of ethidium bromide (5 ~1 50 @g/ml). The excitation (360 nm) and emission wavelengths (590 nm) of ethidium bromide and its complex with nucleic acids (5) were isolated using Corning No. 5860 as primary filter and Jena No. OG 1 as secondary “cut-off” filter. The fluorescence can be measured within 1 min after addition and remains stable for at least 3 hr. Blanks and standards (lo-100 ng DNA) were included in the whole procedure. Ethidium bromide and DNA standard were dissolved in phosphate-buffered saline. A DNA stock solution (23 mg/ml) was prepared in 0.1 M ammonia and kept at 4°C. Determination of RNA. If both RNA and DNA content had to be measured, tissue samples were incubated as described above but also without addition of RNAase. RNA standards (20-300 ng) could be introduced, but the alternative for RNA standards, i.e., the use of DNA standards only and using a factor 0.46 as quotient of RNA/DNA fluorescence (c$ ref. 3 and 5), indeed appeared to give similar results. A RNA stock solution (23 mglml) was prepared in phosphate-buffered saline and kept at 4°C. Freeze-drying. In order to examine the effect of freeze-drying, volumes of 20 ~1 homogenate were lyophilized (0.03 mmHg for 5 hr at -5°C) and redissolved in the same volume of water. DNA content was determined according to the present method as well as by following the procedures of Kissane and Robins (4).
ETHIDIUM
BROMIDE
ASSAY
OF
DNA
AND
227
RNA
Statistics. The significance of differences between the values was calculated using the Student t test. A value for p < 0.05 was considered to be statistically significant. RESULTS
Proportionality. Using 2.5 @g/ml ethidium bromide, the fluorescence shows a linear relationship with the DNA content between 2.5 and 100 ng in the 100~~51final volume. The enhanced fluorescence of 2.5 ng DNA differs significantly (10%) from the ethidium bromide blank (0.025 < p < 0.05; II = 6). For RNA the same appears to be valuable between 5 and 200 ng, i.e., twice the amount of DNA. The recovery of DNA-ethidium bromide fluorescence for contents between 10 and 200 ng appeared to be completely after pronase (0.4 mglmi) treatment (102%; SEM 1.6%; n = 10). Pronase treatmerzt. In order to increase the sensitivity of the ethidium bromide DNA assay, the high background fluorescence due to the tissue (see below), pronase and RNAase had to be suppressed. Reduction of pronase and RNAase background fluorescence was obtained by incubation in a small volume which can be diluted five times prior to background fluorescence reading. Different pronase concentrations were tested for their induction of ethidium bromide fluorescence of tissue samples. For homogenate (Table 1) as well as for pieces of frozen-dried sections of the rat neurohypophysis a broad optimum of fluorescence was found between 0.1 and TABLE ENHANCEMENT
OF ETHIDIUM LOBE
BROMIDE
HOMOGENATE
1 FLUORESCENCE
OF RAT
NEURAL
BY PRONASE”
Pronase
concentration
Fluorescenceh
(mgimlf 0
0.004 0.01 0.04 0.1 0.2 0.4 4 8 20
f%‘o)
100 +- 5.4 99 160 208 22.5 230 223 141 110 68
it 6.9 2 4.9 4 4.9 t 4.1 + 6.6 zt 2.3 2.z 4.8 it 3.6 rt 5.2
n Procedure as described in Material and Methods section using about 3Opg fresh weight of rat neural lobe tissue. b Values are given in percentage (?SEM) of the total fluorescence without pronase treatment and are the mean of three determinations.
228
G. J. BOER TABLE
INFLUENCE WITH
2
OF RNAase IN THE PRESENCE OF PRONASE ON THE FLUORESCENCE ETHIDIIJM BROMIDE OF RNA AND DNA ADDED TO A SAMPLE OF RAT NEURAL LOBE HOMOGENATE”
Fluorescence (%‘oy’
Homogenate +200 ng RNA + 100 ng DNA
+ RNAase
-RNAase
100 f 3.6 102 + 7.2 312 ” 11
369 k 13 382 k 20
163 r 7.2
a Procedure as described in Material and Methods section using about 30 Fg fresh weight rat neural lobe and 5 pg/ml ethidium bromide for final fluorescence reading. b Values are given in percentage (?SEM) of the DNA-ethidium bromide fluorescence of the homogenate and are the mean of three determinations.
0.4 mglml (0.5 < p < 0.6) after 30-min incubation at 37°C. Below 0.1 mg/ml the maximum was not reached; at 20 mglml the fluorescence became even lower than without incubation. The concentration of 0.1 mg/ml pronase was chosen in the final procedure, giving an autofluorescence of about 6% of the ethidium bromide blank fluorescence. Incubation with nucleuses. In the presence of pronase RNAase should break down all RNA present in the tissue sample without attacking DNA. In our procedure the fluorescence of 200 ng RNA, added to a homogenate sample, disappeared completely with the RNAase concentration used (50 pglml), while the fluorescence due to 100 ng DNA added remained present (Table 2). Both nucleic acids were added in amounts that were expected to be present in the 30 pg of tissue maximally used for the micro DNA assay. DNAase (50 pglml) degrades 100 ng DNA totally, independent of the presence of homogenate and pronase. Since no residual fluorescence was observed, DNAase treatment had not to be incorporated in the microassay. Linearity with tissue samples. The linear relation between ethidium bromide fluorescence and sample size is shown for liver homogenate in Fig. 1. Since this linearity is also present in the absence of RNAase, both DNA and RNA can be determined. Background JEuorescence. The background fluorescence for liver is very high (about the value of the enhancement in fluorescence caused by the DNA in the sample), which reduces the sensitivity of the assay seriously (see Discussion). Nevertheless as little as 5 pg wet weight of liver tissue appeared to be enough to give reproducible fluorescence increases. For this weight (the lowest value of Fig. 1) the enhanced fluorescence for the DNA present is about 10% of the total fluorescence. This differs significantly (20%) from the ethidium bromide blank
ETHIDIUM
BROMIDE
229
ASSAY OF DNA AND RNA
FIG. 1. Linear relationship between the amount of rat liver tissue and the enhancement of the ethidium bromide fluorescence by its DNA content (lower line) and by its total nucleic acid content (upper line). Each point indicates the mean of three determinations.
fluorescence (0.025 < p =C0.05; n = 3) and corresponds to about 11 ng of DNA. The background fluorescence of liver expressed per unit of wet weight was about 5 times as high as for the neurohypophysis. For both liver (Fig. 1) and neurohypophysis the tissue background fluorescence did not interfere with the linearity of fluorescence and sample size. Freeze-drying of the homogenates had no apparent effect on the DNA TABLE DNA
3
AND RNA CONTENT OF RAT NEURAL LOBE AND DIFFERENT PRETREATMENTS OF THE TISSUE Two DIFFERENT MICROASSAYS
LIVER MEASURED AFTER AND USING
DNA content? Present micromethod Neural lobe Homogenateb Fresh Frozen-dried
1.16 t 0.04 1.22 2 0.07
Frozen-dried section samples* 9.48 k 1.3
Liver 2.76 k 0.09 2.48 2 0.14 13.3 + 0.9
Method of Kissane and Robins Neural lobe 1.08 ? 0.05 1.30 k 0.20 10.9 + 0.7
Liver 2.41 t 0.27 2.67 5 0.14 14.6 + 1.0
RNA content” Neural lobe
Liver
1.84 n.d.c
6.52 n.d.
12.5
44.4
a Values expressed as ngIpg fresh weight and ngipg dry weight for homogenates and section samples, respectively; each value is the mean (SEM) of three determinations for the homogenates and of five for the section samples. * Homogenate of neurohypophysis originates from three animals from which one was used for preparing liver homogenate; frozen-dried section samples originate from one single animal. C Not determined.
230
G. J. BOER
content measured with the ethidium bromide procedure as well as with the diaminobenzoic acid procedure (Table 3; p-values always above 0.30). Comparison with the diaminobenzoic acid method. After carrying out all pretreatments used for rat neurohypophysis and liver, both the present microassay and the assay with diaminobenzoic acid resulted in similar data for the DNA content (Table 3; p-values always above 0.20). The tissue background fluorescence of the present assay, however, lowered the sensitivity of DNA measurement in such a way that three times more tissue DNA is necessary to give the same increment in fluorescence compared to the blank as with the diaminobenzoic acid method (4). DISCUSSION
The proportionality of DNA and RNA with their ethidium bromide fluorescence in the nanogram range and the nucleotide differentiation using specific nucleases (Table 2) show that the DNA and RNA measurement with ethidium bromide as described by Le Pecq and Paoletti (5) is applicable also at the microlevel. As little as 2.5 ng DNA or 5 ng RNA in solution can be reproducibly detected. Karsten and Wollenberger (3) introduced pronase treatment of tissue homogenate and surface fluorometry which also allows a quantitative reaction of tissue nucleic acids with ethidium bromide. Pronase treatment was also applied within the microprocedure, although sonication of tissue homogenate might be considered as an alternative (1). A higher pronase concentration, however, appeared to be necessary for maximal increase of the fluorescence. For the DNA assay also the RNAase activity had to be raised in order to break down all the RNA expected. Normal fluorometry instead of surface fluorometry could be applied in the present microassay because a linear relationship exists between RNA- and DNA-ethidium bromide fluorescence and sample size with this simpler procedure (Fig. 1). If tissue DNA and RNA content were measured with the present microassay, the sensitivity limit was lowered two to three times compared to the assay for dissolved nucleic acids, mainly by the tissue-induced background fluorescence (which, moreover, appeared to be dependent on the origin of the tissue). The minimum of handling, however, makes the ethidium bromide method favorable if more than about 10 ng of native DNA is available for an assay. Moreover, the ethidium bromide method also introduces a micro RNA determination in small tissue samples. Results of DNA estimations in rat neural lobe and liver homogenates show good agreement with data obtained following the Kissane and Robins (4) microprocedure (Table 3, first line) and are fully comparable
ETHIDIUM
BROMIDE
ASSAY OF DNA AND RNA
231
with available literature values (3,9). Also the presented RNA values are fully in agreement with the data of these authors. Freezing and thawing of tissue did not influence nucleic acid-enhanced ethidium bromide fluorescence (3), while also freeze-drying of tissue homogenates had no influence on the DNA assay (Table 3, second line). Therefore, the present method seems also valuable for small frozendried tissue samples as are used in microbiochemistry (7). Using such samples prepared from rat neural lobe and liver, indeed, with both micromethods similar data were obtained for the DNA content (Table 3, third line). The reproducibility of the present microassay lies in the same order as for the diaminobenzoic acid method (Tables 2 and 3). The present method has already been used for several months for quantitative determination of the accumulation of neurohypophyseal glial cells (pituicytes) of the rat along different micro density gradients (Ficoll and Ludox) after centrifugation. Good recoveries were obtained. If necessary, the procedure can also be scaled up two or three times if more sample is available, thus avoiding the errors inherent in handling small volumes. ACKNOWLEDGMENTS I am grateful to Dr. D. F. Swaab for his critical remarks in the present study, and thank Mr. B. Fisser, Miss A. Nolten, Mrs. C. van Rheenen. and Miss A. van der Velden for their successive assistance and Prof. Dr. J. Ariens Kappers and Prof. Dr. J. M. Tager for their revision of the manuscript.
REFERENCES 1. Blackburn. M. J., Andrews. T. M., and Watts, R. W. E. (1973) Anal. Biochem. 51, l-10. 2. Boer, G. J., and Jongkind, J. F. (1974) J. Neurochem. 22, 965-970. 3. Karsten. U., and Wollenberger, A. (1972) Amd. Biochem. 46, 135-148. 4. Kissane, J. M., and Robins, E. ( 1958) J. Biol. Chem. 233, 184- 188. 5. Le Pecq. J.-B., and Paoletti, C. (1966) And. Biochem. 17, 100-107. 6. Le Pecq, J.-B., and Paoletti, C. (1967) J. Mol. Biol. 27, 87-106. 7. Lowry, 0. H. (1953) J. Histochem. Cytochem. 1, 420-428. 8. Neuhoff. V. (1973) Micromethods in Molecular Biology, p. 399. Springer-Verlag, Berlin. 9. Sachs, H., Saito, S., and Sunde, D. (1971) in Memoirs of the Society for Endocrinology (Heller. H., and Lederis, K., eds.), p. 325. Cambridge Univ. Press, London,
